Text

Introduction: Workers in occupations that require prolonged standing, e.g., nurses or hairdressers, experience increased risks of developing varicose vein (VV). Most recent evidence-based guidelines published by the American Venous Forum recommend non-invasive Duplex ultrasound scanning (DUS) to visualize the vein valve function for standardized diagnoses of VV. But, reproducibility of DUS, especially when applied in epidemiological field studies, remain unclear. Knowledge about the expected normal range of measurement variability or how results differ when two measurement teams examine the same individual is limited. The aim of this study was to determine inter-rater, test-retest, and inter-observer reliability of DUS for describing an individual’s lower extremity chronic venous disorder (CVD) status.

Methods: Thirty-three participants (20♀) were examined via DUS on seven vein segments of the same leg on two days by two measurement teams (two observers each), considering recommendations for standardization (morning, standing position, manual reflux-provocation manoeuvre). Test-retest reliability involved one rater evaluating DUS images of the same individual taken by the same team on both days. Inter-rater reliability was assessed by having two raters evaluate the same image, while inter-observer reliability was determined by having one rater evaluate the images performed by both teams the same day. Bland-Altman plots (incl. systematic bias), intraclass correlation coefficient (ICC), and standard error of measurement (sem) were computed. Regarding the categorical assessment of pathologic reflux, the cutoff value >500 ms is defined for superficial and >1000 ms for deep veins.

Results: One participant showed pathological reflux in three great saphenous vein segments from groin to calf; three others in a single segment. Overall, our sample showed rather healthy vein status. No systematic errors were found for DUS in all vein segments. Inter-rater reliability in DUS was very high (sem-range=[0.05 to 0.13]; ICC-range=[0.45 to 1.00]), compared to test-retest (sem-range=[0.07 to 0.93]; ICC-range=[-0.21 to 0.94]) and inter-observer reliability (sem-range=[0.12 to 0.61]; ICC-range=[-0.01 to 0.95]). A high absolute agreement (≥94%) across all vein segments was shown for the pathological reflux classification.

Discussion: Results indicate that a high reproducibility of DUS is achievable when considering recommendations for standardization, although if well-trained but not medically trained observers carry out the assessments. Although reliability parameters were partly contradictory, categorical agreement – clinically more relevant – was very high. Given the rather vein-healthy sample, DUS reliability should be evaluated in more heterogenic samples in future studies.

Conclusion: Findings help to facilitate CVD diagnostics, e.g., when planning future intervention studies or longitudinal field studies investigating standing as a risk factor for CVD. A high degree of objectivity can be expected even though rater and observer do not necessarily need to be the same person. Using several raters evaluating the same DUS images and finding consensus in case of disagreement seems reasonable.